Water utilities detect leaks in potable water distribution systems through employing various types of sensors, mainly pressure sensors or hydrophones, and analytical methods such as correlation and spectral analysis in their networks. A fusion of these various data sources and analytical models can help towards achieving a more refined, accurate, and complete estimates and assessments. In this research, we approach the leak detection from a data and information fusion perspective where three types of sensors_ accelerometer, hydrophone, and pressure transducer_ and three supervised classification algorithms_logistic regression, support vector machine, and decision tree_ are implemented. Moreover, the effects of system’s characteristics, like network architecture, leak types, water velocity, and pressure on detection methods are evaluated. Finally, a set of sensors and classification algorithms for accurate leak detection in different systems are recommended.

Application of Supervised Machine Learning Techniques to Detect Leakage in a Water Distribution Testbed Using Accelerometers, Hydrophones, and Pressure Transducers

Created in 1999, the Environmental & Water Resources Institute (EWRI) is the recognized leader within ASCE for the integration of technical expertise and public policy in the planning, design, construction, and operation of environmentally sound and sustainable infrastructure impacting air, land and water resources.

[Paper Proceedings](https://doi.org/10.1061/9780784483466)

Conference DOI: https://doi.org/10.26226/m.6329c39b68fb0e00131f77d2